ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent/Changes
(Generate patch)

Comparing AnyEvent/Changes (file contents):
Revision 1.232 by root, Thu Apr 23 22:44:30 2009 UTC vs.
Revision 1.634 by root, Sun May 12 11:14:13 2019 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: perl idle 3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
10TODO: maybe implement env variable to give hosts precedence
11TODO: more formats for parse_ipv6 (single ipv6 address without port, ...p80, ...#80)
12TODO: inet_aton maybe pack C4? no, add a comment why not
13TODO: socket prepare callbacks, can they easily signal failure as well?
14
15TODO: Олег Г <verdrehung@gmail.com>, empty A record response causes
16AnyEvent::resolve_sockaddr to eventually call pack_sockaddr with undef -
17broken dns server, but maybe improve the response.
18
19 - AnyEvent::Log did not reassess logging status of
20 AnyEvent::Log::logger's when contexts were changed with
21 ->attach/detach/slaves, causing them to not log even though a
22 recent attach should have caused them to log.
23 - added some more logging configuration examples.
24
257.15 Tue Feb 26 03:07:42 CET 2019
26 - INCOMPATIBLE CHANGE: AnyEvent::Handle's tls_detect documentation gave
27 separate major and minor versions, while code passed only a single
28 value. This version follows the documentation and now passes separate
29 major and minor values.
30 - work around Net::SSLeay not having been ported to openssl 1.1, but many
31 distributions compiling it against openssl 1.1, which unfortunately
32 succeeds and results in a very broken module.
33 - AnyEvent::DNS::dns_unpack now stores the original DNS packet
34 in the __ member, to allow decoding of undecodable resouce records
35 containing compressed domian names.
36 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
37 addresses. It now does and converts them to ipv4 mapped addresses.
38 - support CAA records, based on patch by Steve Atkins.
39 - add freenom and cloudflare nameservers as dns fallback.
40 - AnyEvent::Strict would not properly ward against io watchers
41 on files when the handle passed was a file descriptor.
42 - document "internal" variables used by the dns en-/decoder to allow
43 enterprising users to extend them in a semi-official way.
44
457.14 Sat Jun 24 01:44:19 CEST 2017
46 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
47 by Paul Howarth and apparently tracked down by Mike McCauley).
48 - AnyEvent::Handle->keepalive was documented (and defined) twice
49 (found by Matt Merhar).
50 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
51 listening socket unless a guard is used. Change this so that no
52 cleanup will be performed unless a guard is used and document this more
53 clearly.
54 - make tcp_bind/tcp_server error messages more regular.
55 - work around latest perl upstream breakage - this time, perl5porters were
56 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
57 stupidly apply anything that has "security fix" stamped on it, it's likely
58 going to be a problem in practise (and working around it is trivial on
59 windows and unix).
60 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
61 to constants.
62 - update warnings to common::sense 3.74 standards.
63
647.13 Sat Sep 17 04:31:49 CEST 2016
65 - Only call tlsext_host_name for non-empty common names (reported
66 by Maxime Soulé).
67 - log a (single) notice message if SNI is not supported.
68 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
69 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
70 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
71 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
72 testcases and the specification and not in the code - the post-9.0.0
73 unicode files have all known problems fixed, so finally the AnyEvent
74 IDNA implementation can pass the full IDNA testsuite - without needing
75 a single fix :)
76 - guarantee (and document) that condvar callbacks will be removed
77 on invocation - important to avoid circular references.
78
797.12 Wed Jan 27 19:12:26 CET 2016
80 - use common name as hostname for tls connects, if Net::SSLeay
81 supports SNI.
82 - fix documentation of tls_autostart read type in AnyEvent::Handle,
83 analyzed by Felix Ostmann.
84
857.11 Thu Jul 16 14:36:00 CEST 2015
86 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
87 addresses (extra "::" at end and similar cases).
88 - add a more explicit warning to AnyEvent::Handle that it doesn't
89 work on files, people keep getting confused.
90 - new function AnyEvent::Socket::tcp_bind.
91 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
92 - aligned ipv6 address formatting with RFC 5952 (by not shortening
93 a single :0: to ::).
94 - added stability canary support.
95
967.09 Sat May 2 16:38:53 CEST 2015
97 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
98 that was renamed to AnyEvent::Log:format_time. uinder its old name
99 (analyzed by Michael Stovenour).
100 - update AnyEvent::DNS fallback resolver addresses:
101 seems google effectively killed most other free dns resolvers,
102 so remove them, but add cable and wireless (ecrc) since it was
103 stable for 20 years or so, official or not, and there should be
104 an alternative to google.
105 - perl5porters broke windows error codes in 5.20, and mapped
106 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
107 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
108 because they were so cool. They probably broke other error codes
109 for no reason, too, but I didn't care for checking, it's too
110 depressing. This version only works around the WSAEWOULDBLOCK
111 issue, because I don't have a nice way to work around the
112 WSAEINPROGRESS bug.
113
1147.08 Wed Dec 10 05:27:17 CET 2014
115 - work around a newly introduced bug in Socket 2.011 (an
116 errornous sun_length check) (analyzed by Maxime Soulé).
117 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
118 (analyzed by Ben Magistro).
119 - AnyEvent::Strict will now confess, not croak. This is in line with
120 it being a development/debugging tool.
121 - work around a number of libglib bugs (debug builds of libglib enforce
122 certain undocumented behaviour patterns such as not being able to
123 remove a child watch source after it has fired, which we will try
124 to emulate to avoid "criticals". what where they thinking...).
125 - mention json security issues in AnyEvent::Handle, now that Douglas
126 Crockford has foolishly and incompatibly changed JSON.
127 - changed default dns resolver "max_outstanding" value from 1 to 10,
128 the latter beinfg the intended value all along
129 (reported by Ilya Chesnokov).
130 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
131 (written by Mike Lowell).
132
1337.07 Tue Dec 17 17:45:02 CET 2013
134 - the documentation for custom tls verify schemes was wrong. make it agree
135 with the code (reported by Maxime Soulé).
136 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
137 - work around an API change in openssl that could cause wrong tls connection
138 aborts, likely on windows only (analyzed by sten).
139 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
140 endless loop (reported by Dietrich Rebmann).
141 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
142 (patch by Maxime Soulé).
143 - document AnyEvent::Impl::IOAsync::set_loop and
144 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
145 functionality has _always_ been available.
146 - force a toplevel domain name in t/81_hosts.t (analyzed by
147 David Jack Wange Olrik).
148 - document that AnyEvent::Log uses AnyEvent::IO.
149 - warn about AnyEvent::Filesys::Notify performance.
150 - praise the joys of AnyEvent::Fork::*.
151 - time for an =encoding directive.
152 - do no longer use JSON to create a default json coder, use
153 JSON::XS or JSON::PP directly.
154
1557.05 Wed Aug 21 10:38:08 CEST 2013
156 - uts46data.pl couldn't be found due to wrong naming of the file
157 (reported by Fulko Hew).
158 - handle lone \015's properly in AE::Handle's default line read
159 (reported by various people).
160 - untaint ip addresses found in /etc/hosts (patch by José Micó).
161 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
162 fail on first use (reported and testcase by Andrew Whatson).
163 - expose AnyEvent::Log::format_time, and allow users to redefine it.
164 - expose AnyEvent::Log::default_format, and allow redefinition.
165 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
166 - AnyEvent::Debug shell can now run coro shell commands, if available.
167 - t/63* tests were wrongly in MANIFEST.
168 - kernel.org's finger server went MIA, switch to freebsd.org and
169 icculus.org.
170 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
171 to function asynchronously (reported by Daniel Carosone).
172 - hard-disable $^W in most tests, it generates too much garbage output.
173 - use a (hopefully) more future-proof method to emulate common::sense.
174 - upgrade to UTS-46:6.2.0.
175 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
176 documented. should not affect anything.
177
1787.04 Thu Nov 15 02:23:58 CET 2012
179 - debugging code left in.
180
1817.03 Thu Nov 15 02:16:11 CET 2012
182 - AnyEvent::Socket::inet_aton did not work when dns resolution was
183 used to find the addresses (analyzed and patch by Maxime Soulé).
184 - fix a memory leak in the /etc/hosts lookup code when hosts don't
185 resolve and are not in hosts (analyzed by sten).
186
1877.02 Tue Aug 14 04:27:58 CEST 2012
188 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
189 program by Yorhel).
190 - verified that AnyEvent::Socket follows RFC5952.
191 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
192
1937.01 Sun May 13 01:03:17 CEST 2012
194 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
195 not available, instead of throwing an exception.
196 - use File::Spec to get the tmpdir in t/*, to avoid needless
197 failures on (most, not mine :) windows boxes.
198 - new handle read types: tls_detect and tls_autostart.
199
2007.0 Fri Apr 13 06:33:30 CEST 2012
201 - child watchers are broken in POE 1.352 (also many earlier
202 versions) and there seems to be no way to work around it, as POE
203 itself is inherently racy. Document this regression and add a delay
204 in t/68_poe_03_child.t for the time being.
205 - new module AnyEvent::IO, that is a frontend to either a pure-perl
206 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
207 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
208 - load /etc/hosts only when DNS has no answer.
209 - stat /etc/hosts on every access and reload it if it changed.
210 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
211 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
212 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
213 a side effect, they now use true appending as opposed to libc
214 appending, and the file might not have been opened when the function
215 returns.
216 - the default logging level was not properly documented in a variety of
217 places, this has hopefully been rectified.
218 - updated uts46data.pl for unicode 6.1.0.
219 - made log messages generated by AnyEvent submodules not
220 include the package name anymore, as it will be logged
221 by default already.
222 - upgrade to the trick used by common-sense 3.5 to work
223 around extra warning torture/breakage under perl 5.15.x.
224 - log messages by anyevent are now ucfirst, are usually full
225 sentences and do no longer include the package name.
226 - the storable read type would sometimes throw an exception instead
227 of causing EBADMSG (analyzed by Maxime Soulé).
228 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
229 keysize (reported by Paul Howarth).
230 - documented Rocco fabricating statements about AnyEvent and me.
231
2326.14 Tue Jan 31 20:00:24 CET 2012
233 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
234 probably an editing mistake (reported by Darin McBride).
235
2366.13 Thu Jan 12 07:27:01 CET 2012
237 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
238 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
239 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
240 regex-matching inside the callback would change the parameter, despite
241 $1 being dynamically scoped per-block. this perl bug is now being
242 worked around (testcase by Cindy Wang).
243
2446.12 Mon Dec 12 13:21:10 CET 2011
245 - $! was clobbered by subsequent calls in tcp_connect, due to
246 the postpone not saving/restoring it (analyzed by Richard Garnish).
247
2486.11 Tue Nov 22 10:36:05 CET 2011
249 - Tk cannot create windows when tainted unless you set a title, so
250 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
251 - escape any nonprintable/nonascii characters when stringifying
252 backtraces.
253 - log the reason loading an interface module fails to load at level debug.
254
2556.1 Tue Oct 4 19:44:30 CEST 2011
256 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
257 and above), and some messages inside AnyEvent have been elevated
258 to higher log levels to print by default.
259 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
260 - work around yet another signal race bug, newly introduced by POE. My
261 how that module sucks.
262 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
263 support ipv6-only operation even with defaults (also use ::1 instead
264 of 127.0.0.1).
265 - give better guidelines on how to chose a logging level.
266 - AnyEvent::Log can now cap (limit) log message priority by context.
267 - AE::log will now save and restore $! and $@, so callers don't have to.
268 - AE::Strict now checks that registered fds are still valid at regular
269 intervals.
270 - skip frequently called 'require' calls if the module appears loaded.
271 - increase signal latency test timer considerably for poor little osx,
272 which can't be bothered with basic posix functionality such as signal
273 delivery.
274 - mention some debugging aids in the tutorial.
275 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
276 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
277 of the context as title.
278 - AnyEvent::Strict would not restore read-write access to $_ when a callback
279 died - this is a user bug, but since AE::Strict exists to help dbeugging...
280 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
281 to work with it - the author is well aware of the issues but refuses
282 to fix them. This extreme step has been taken because the module actively
283 attacks the goals of AnyEvent and makes it harder for module authors to use
284 AnyEvent.
285
2866.02 Fri Aug 26 20:08:31 CEST 2011
287 - log_to_syslog now accepts facility strings.
288 - log_to_syslog would not allow facilities - "$facility|$level"
289 *literally* means that string. sheesh.
290 - the CPAN parser got confused and thought we provide Sys::Syslog, try
291 to work around.
292 - renamed AnyEvent::Impl::FLTK2 to FLTK.
293
2946.01 Fri Aug 26 07:04:11 CEST 2011
295 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
296 the TTL as fourth element - this affects ->resolve, ->request
297 and ->dns_unpack indirectly. Convenience functions such as
298 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
299 AnyEvent::Socket.
300 - INCOMPATIBLE CHANGE: backend authors now should not implement
301 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
302 methods.
303 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
304 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
305 - DEPRECIATION: please change
306 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
307 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
308 in your programs, for future compatibility.
309 - *blush* idle watcher emulation was borked.
310 - AnyEvent::Strict now checks the AE::xxx functions as well.
311 - querying too long domain names would barf deep inside AnyEvent::DNS - since
312 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
313 for too long domain names explicitly and passes an undef to the callback
314 instead (reported by James Bromberger).
315 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
316 - new functions: AnyEvent::postpone, AnyEvent::log.
317 - new module: AnyEvent::Log, for simple logging needs.
318 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
319 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
320 a debug shell.
321 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
322 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
323 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
324 /etc/hosts can be found.
325 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
326 configuration.
327 - env variables can now be specified using AE_xyz instead of
328 PERL_ANYEVENT_xyz, subject to some rules.
329 - lots of new AnyEvent::Debug functionality.
330 - AnyEvent::Strict now checks for modifications of $_ by the callback
331 (by setting it to readonly while executing the callback, if possible).
332 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
333 Invest extra overhead to work around this peculiar behaviour.
334 - split the pure perl backend into a loop module and an impl
335 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
336 - some read types (regex, netstring, json, storable) could
337 errornously keep an AE::Handle object alive on their own
338 (reported by Mohammad Toossi).
339 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
340 supports /etc/hosts overrides.
341 - add more workarounds around child watcher bugs in IO::Async
342 and (newly introduced) in POE.
343 - use glib child watchers and try to work around its limitations,
344 if possible (based on a patch by Kevin Ryde).
345 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
346 - many read types in AnyEvent::Handle would malfunction during
347 transitions to TLS because of a "delete $self->{rbuf}".
348 (reported by Gerald Galster).
349 - improve condvar blocking wait performance (probably only noticable
350 with faster backends), also make it easier for event loops that
351 do not support blocking waits.
352 - PERL_ANYEVENT_MODEL now also allows full module names, although
353 the usefulness of this feature is not yet clear.
354 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
355 - add t/80_ssltext.t.
356 - no longer use AUTOLOAD for the initial loading (for no very good reason).
357 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
358 - do not run event loop detection multiple times when programs call their cached
359 copies of AnyEvent::detect.
360 - parse_hostport now accepts absolute paths als unix domain sockets.
361 - suppress the idiotic warning inside POE by patching the POE::Kernel
362 run flags instead of asking the user to jump through contortions.
363 - add Task::Weaken as dependency (as recommends only at this time).
364 - replace PF_UNSPEC by 0, for good old minix.
365
3666.0 (not officially released, but was on CPAN for about 15 minutes :/)
367
3685.34 Thu May 12 09:59:41 CEST 2011
369 - AE::TLS didn't support both cert_file and key at the same time,
370 due to some copy&paste bug (reported and analyzed by vti).
371 - improve AE::Handle timeout documentation slightly.
372
3735.33 Mon Feb 7 21:51:41 CET 2011
374 - never appeared on CPAN, for unknown reasons.
375 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
376 object when it's already gone (analyzed by Ryan Bullock).
377 - update idna algorithm(s) to UTS#46 version 6.0.0.
378 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
379 uts46 says it's an error, and testsuite makes more sense.
380 - remove leading dots in idn_to_ascii - testsuite says to remove,
381 uts46 says it's an error, and testsuite makes more sense.
382 - fixed a bug in idn ignored character processing.
383 - passes the uts46 6.0.0 testsuite except for three cases which
384 apparently are bugs in the testsuite itself.
385 - passes the uts46 6.0.1 testsuite except for eleven cases which
386 apparently are bugs in the testsuite itself.
387
3885.31 Sun Jan 23 11:49:19 CET 2011
389 - work around a POE bug where POE would first poll for new events and
390 invoke handlers BEFORE handling the message queue.
391 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
392 - stop_read now does just that in TLS mode. see its
393 documentation for caveats.
394 - update to common::sense 3.4 warning set (some warnings in perl impose
395 an arbitrary 32kb limit on regex matches, which affects
396 AnyEvent::Handle).
397 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
398 - new experimental AE::Handle::resettls method.
399
4005.3 Fri Dec 31 05:48:13 CET 2010
401 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
402 issues with IO::Async have been solved in that version, including
403 the availabilty of a default loop!
404 - implement new wbuf_max parameter for AnyEvent::Handle.
405 - added Cocoa::EventLoop backend by Daisuke Murase.
406 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
407 - clarify handle on_prepare documentation.
408 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
409 to avoid leaking them to any child processes.
410 - try to force files into the arch-dependend lib directory, for the
411 half-"turly obsessive compulsive person"(s) that depend on it.
412
4135.29 Sun Dec 5 10:49:21 CET 2010
414 - convert EV backend to EV 4.00 API (so better upgrade EV too).
415 - AE::Handle: implement dynamic read_size adjustment, add
416 max_read_size parameter, reduce default read_size to 2048.
417 - add some FAQ entries for servers and/or on_eof vs. on_error.
418 - work around OS bugs (cygwin again) when lingering in AE::Handle.
419 - reduce memory usage slightly when connect was used in AE::Handle.
420 - wrok around more segfaults in Tk.
421 - document tls_ctx => undef as valid.
422 - detect WNOHANG value at compiletime.
423 - include some probably linux-only support for building AnyEvent
424 as part of the perl core.
425 - improve compatibility of t/08_idna.t to perl 5.8.
426 - make AnyEvent::Strict truly optional (in case somebody builds a
427 minimal perl...).
428
4295.28 Wed Oct 13 04:14:23 CEST 2010
430 - due to a glitch, AnyEvent's internal getprotobyname wasn't
431 actually used in tcp_server.
432 - implement AnyEvent::CondVar->new method, which was documented but
433 missing.
434 - support autoloading of read types in unshift_read.
435 - AnyEvent::Strict: do the test for invalid fh arguments differently,
436 resulting in more informative error messages.
437 - new AnyEvent::FAQ manpage.
438 - clarify that you can actually call push_read etc. in on_read.
439 - put keys and certs in /tmp, not into the current directory,
440 and document this in the SECURITY CONSIDERATIONS sections
441 of AnyEvent::TLS.
442 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
443 requested by Ron Isaacson.
444 - document $AnyEvent::DNS::RESOLVER variable.
445 - incorporated major typo patches by Abhijit Menon-Sen.
446 - AnyEvent::Handle now croaks when negative timeouts are passed.
447 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
448 undocumented, but who knows when they might come in handy.
449
4505.271 Tue Jun 8 12:05:46 CEST 2010
451 - backport to perl 5.8.x.
452
4535.27 Sun Jun 6 12:12:05 CEST 2010
454 - postpone differently in AnyEvent::Socket now, as
455 when not, canceling the connection attempt might fail
456 (found by Felix Antonius Wilhelm Ostmann).
457 - explicitly check for non-stream sockets in AE::Handle, too many
458 clueless people fell into the trap of this somehow working.
459 - simplified and reworked the "OTHER MODULES" section.
460 - better/more condvar examples.
461
4625.261 Wed Apr 28 16:13:36 CEST 2010
463 - AF_INET6 was not properly used from Socket6 during configuration
464 time (found by Andy Grundman).
465
4665.26 Mon Apr 12 04:49:35 CEST 2010
467 - don't generate spurious readyness notifications when select
468 returns EINTR in the pure perl backend (analysed and testcase
469 by Bas Denissen).
470 - give same examples in SYNOPSIS sections of AnyEvent and AE.
471 - provider faster implementations for AE::time/now/now_update for
472 EV and Perl backends.
473 - provide faster AE::cv and AE::time implementations for all
474 backends.
475 - no longer support lower/mixed-case signal names (this was
476 never documented nor universally supported).
477 - some more memory µ-optimisations, and somewhat less messy
478 code to cope with both AnyEvent and AE APIs.
479
4805.251 Sat Mar 13 00:58:21 CET 2010
481 - make test failed if it was run before make install on perl <=
482 5.8.8, and older versions of perl make our live very difficult,
483 so write constants.pl during Makefile.PL time.
484
4855.25 Sat Mar 13 00:23:14 CET 2010
486 - fix a race condition in AnyEvent::Handle that would cause
487 a "bio_read: unsupported method" error (in _another_
488 ssl connection) after user code threw an exception.
489 - added AnyEvent::Handle->destroyed method.
490 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
491 - the AnyEvent::Util::fh_nonblocking function worked only
492 by lucky accident on win32.
493 - smaller and faster AnyEvent::Util::fh_nonblocking.
494 - when the (required!) Time::HiRes module is missing, AnyEvent
495 did not fall back to built-in time properly.
496 - do not load Fcntl at runtime, saving memory and loading time.
497 - precompile a number of constants and use them instead of runtime
498 detection and eval.
499 - free detection code after detection and similar memory optimisations.
500 - Perl backend timer interval best effort drift has been improved
501 (same algorithm as EV).
502 - update unicode idna mapping table.
503
5045.24 Tue Jan 5 11:39:43 CET 2010
505 - cygwin never reports errors from failed connects. 1.5 just gives
506 you continous readyness and EAGAIN, 1.7 is even more broken and
507 just hangs. work around both issues in a cygwin-specific hack.
508 - improve idle watcher documentation slightly.
509
5105.23 Sun Dec 20 23:48:00 CET 2009
511 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
512 - implement punycode_encode/decode, idn_nameprep,
513 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
514 - provide $AE::VERSION.
515 - removed traces of "no strict 'refs'".
516
5175.22 Sat Dec 5 03:51:13 CET 2009
518 - downgrade-or-fail in AnyEvent::Handle::push_write, to
519 diagnose encoding failures earlier and more succinctly.
520 (this works around bugs in perl, throwing away encoding info
521 when passing scalar data to extensions).
522 - add more examples to AnyEvent::Socket manpage.
523 - upgrade internal warning set to the same as common::sense 2.03.
524 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
525 (John Beppu).
526 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
527 - add google public dns servers to fallback server set and make sure
528 we load-balance properly between the three sets. also add all
529 fallback dns servers, not just a random one, to each dns config.
530
5315.21 Thu Nov 19 02:48:47 CET 2009
532 - fix a problem where socket constants were called with parameters
533 (spotted by David Friedland).
534 - fork_call never use'd POSIX (reported by Daisuke Maki).
535 - improve perl 5.6 compatibility further (but it still won't work
536 unless you rip out everything but the core).
537 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
538 uses 10MB of RAM, but doesn't open a console window. *sigh*.
539
5405.202 Wed Oct 14 22:35:44 CEST 2009
541 - AnyEvent::DNS would unexpectedly clobber $_ under windows
542 (analysed by Matthias Waldorf).
543 - AnyEvent::Handle::run_cmd can now pass the PID of the
544 newly-created process, which is much less useful than it might
545 sound (based on patch by Yann Kerherve).
546
5475.201 Tue Sep 29 12:09:25 CEST 2009
548 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
549 (reported by Torsten Foertsch).
550 - common::sense 2.0 could cause tcp_server to throw an exception
551 (analysed by elmex).
552
5535.2 Mon Sep 14 07:04:49 CEST 2009
554 - INCOMPATIBLE CHANGE: do no longer support register_read_type
555 and register_write_type in AnyEvent::Handle, instead support
556 package names (the facility was mostly abused).
557 - implement "packagename-as-read/write type" support in
558 AnyEvent::Handle.
559 - AnyEvent::Handle: new options "keepalive" and "oobinline".
560 - oobinline set by default to avoid security issues.
561 - the pure-perl event loop backend wrongly detected times() fallback
562 support (spotted by Pavel Boldin).
563
5645.12 Tue Sep 1 20:26:50 CEST 2009
565 - be more lenient when parsing resolv.conf files, as some people
566 use hashmarks as comment indicator inside directives (reported
567 by Michael S. Fischer).
568 - use same set of warnings as common::sense 2.0.
569 - fix a potential 32 bit overflow issue due to perl having problems
570 with large hex constants in the Perl backend.
571 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
572 by default.
573 - fix some prototypes.
574 - (optionally) require Async::Interrupt 1.02.
575 - replace opendns fallbacks by verizon/level3 public dns servers.
576
5775.112 Fri Aug 21 13:59:04 CEST 2009
578 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
579 using starttls with an illegal/uncreatable tls context.
580 - port to broken win32 perls.
581
5825.111 Sun Aug 16 18:44:36 CEST 2009
583 - perl errornously requires sockaddr_un structures returned by the
584 kernel to have a specific length. work around this bug by padding
585 sockaddr_un structures (found to fail on OS X, as perl only uses
586 the correct code for GNU/Linux, but likely to fail on other
587 OSes as well).
588
5895.11 Wed Aug 12 17:49:37 CEST 2009
590 - fixed a bug in run_cmd with input from filename.
591 - create a json coder object when none is given in write direction
592 to reduce dependency on JSON.
593
5945.1 Tue Aug 11 03:17:32 CEST 2009
595 - tcp_connect now reports the correct error in $!, not an unrelated
596 one (was broken since 4.91).
597 - AE::Handle did not properly clear rtimeout/wtimeout watchers
598 when retrying a tcp_connect.
599 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
600 - fixed wrong documentation in AE::Socket.
601
6025.01 Mon Aug 10 03:16:32 CEST 2009
603 - last release broke EV child watchers :/.
604
6055.0 Sun Aug 9 17:32:09 CEST 2009
606 - officially introduce and document the AE API.
607 - lowering the timeout at runtime did not have immediate effect
608 in AnyEvent::Handle.
609 - AE::Handle now has separate and independent read and write timeout
610 settings.
611 - probe for Guard module when AnyEvent::Util is loaded, not at
612 runtime (Event doesn't like eval "use" from callbacks, and it avoids
613 exporting the probe function).
614 - Impl::IOAsync's io watchers did not use the filehandle emulation
615 code to work around it's watcher limits.
616 - work around IO::Async's problems with overloaded objects.
617 - try to disable wearnings inside IO::Async, it's just too buggy.
618 - do no longer use Event by default, only EV or Perl, Event is too buggy.
619 - hackishly provide fast AE:: interface when EV or Perl is the backend.
620 - inherit VERSION from AnyEvent::VERSION in most modules.
621
6224.91 Thu Aug 6 15:42:45 CEST 2009
623 - AE::Handle::starttls could get out of sync when the read buffer
624 already contains some TLS handshake.
625 - AE::Handle did not properly free the TLS session, causing bigger
626 memory leaks in Net::SSLeay than expected (reported by toaster).
627 - AE::Socket::tcp_connect will now artificially delay invoking
628 the callback to avoid returning after invoking the callback.
629 - convert many internal watcher uses to AE API.
630
6314.9 Sat Aug 1 11:07:01 CEST 2009
632 - Glib silently fails when registering a timeout with
633 a negative value, so avoid that.
634 - call condvar callback immediately when it is set
635 after the condition is already signalled.
636 - check rbuf_max condition only after trying to consume
637 data.
638 - normalise signal numbers to names when using the ae's
639 signal handling, but do not document this (yet).
640 - pure perl signal emulation did not properly set nonblocking
641 mode on the signal pipe on win32 (this is unlikely to have caused
642 any issues).
643 - new module: AnyEvent::Debug.
644 - AnyEvent::Strict now checks that a signal was specified
645 by name (not name or number).
646 - reduce memory footprint in typical cases by ~50kb by
647 only compiling signal code when necessary.
648 - add AnyEvent::Handle->rbuf_max.
649 - grab the AE namespace for future new API, implement stubs for
650 said future/alternative API.
651 - new function: AnyEvent::Socket::format_hostport.
652
6534.881 Tue Jul 28 12:51:53 CEST 2009
654 - work around a bug in local in pre-5.10 perls, causing
655 AnyEvent::Handle to recurse when it should not
656 (analysed by elmex).
657
6584.88 Tue Jul 28 04:04:37 CEST 2009
659 - re-bless the handle into a dummy package after calling
660 AnyEvent::Handle::destroy, so the user does not need to check
661 for errors after every push_write etc.
662 - do not attempt to run t/02_signals.t on obviously broken
663 platforms.
664
6654.87 Sun Jul 26 02:06:16 CEST 2009
666 - do not attempt to linger when there is no longer a valid fh.
667 - enforce tls mode to be either accept or connect, do not simply
668 segfault in Net::SSLeay.
669 - AnyEvent::Handle can now call tcp_connect itself (new parameters
670 connect, on_prepare, on_connect and on_connect_error). Updated
671 tutorial accordingly.
672 - add AnyEvent::Impl::Irssi backend.
673
6744.86 Mon Jul 20 23:52:29 CEST 2009
675 - since the verbose warning is not enough, explicitly document
676 that versions before 1.33 of Net::SSLeay are not secure.
677 - work around signal handling races in Event and (...) Event::Lib.
678 - try to align signal-race timer to full-second boundaries.
679 - work around Tk not liking negative timeouts.
680 - don't complain of different grades of environmental unfriendlyness
681 in IO::Async.
682
6834.85 Sat Jul 18 06:16:14 CEST 2009
684 - nail the signal race problem in perl once and for all
685 (see $AnyEvent::MAX_SIGNAL_LATENCY).
686 - take advantage of Async::Interrupt if it is available.
687 - load Time::HiRes and Guard modules on demand only.
688 - add optional/recommended modules section to AnyEvent
689 documentation.
690 - reduce memory usage considerably (and reduce startup penalty)
691 by not using "strict", "warnings" and "overload" modules.
692 - work around buggy windows/openbsd perls and provide EBADMSG
693 and EPROTO ourselves when missing.
694 - improve perl 5.6 compatibility of the core event loop.
695 - made Net::SSLeay version 1.33 a soft requirement.
696
6974.83 Fri Jul 17 16:56:26 CEST 2009
698 - implement AnyEvent::Socket::getprotobyname.
699 - AnyEvent::CondVar's will now detect recursive blocking
700 waits and will croak, as too many people fall into
701 this trap.
702 - AnyEvent::Handle will now call ->destroy on itself after
703 executing the on_error callback, instead of doing some
704 half-baked internal shutdown, for fatal errors.
705 - clarify on_eof behaviour w.r.t. the read queue and
706 on_read callbacks.
707 - ignore some possible spurious wake-ups in tcp_connect.
708
7094.82 Sat Jul 11 00:34:55 CEST 2009
710 - POE and Event backends didn't accept some callable objects as
711 callbacks.
712 - use Config module instead of POSIX module to detect signal names
713 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
714 module doesn't even have all POSIX signals :/.
715 - use more workarounds around the many refcnt/corruption bugs in
716 Event::Lib.
717 - work around a race condition in perl's select, causing t/03_child.t
718 to rarely fail.
719
7204.81 Thu Jul 9 10:30:30 CEST 2009
721 - AnyEvent::Handle didn't properly diagnose write errors
722 (it expected -1 from syswrite, how lame... :).
723 - support file descriptors in addition to file handles
724 in AnyEvent->io.
725 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
726 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
727 and PERL_ANYEVENT_CA_PATH.
728 - provide a sensible synopsis section for AnyEvent::TLS.
729 - add a "supported backends" section to the manpage.
730 - added simple io watcher test to testsuite, using a
731 portable_socketpair.
732 - tried to improve the stability of the Event::Lib backend,
733 YMMV.
734
7354.8 Mon Jul 6 23:45:16 CEST 2009
736 - AnyEvent::DNS did not properly follow CNAME records with
737 uppercase targets.
738 - AnyEvent::DNS would errornously return AAAA records
739 with v4 mapped addresses (a faulty record) as ipv4 addresses,
740 causing AnyEvent::Socket to throw an exception.
741 - added new module AnyEvent::TLS for easier SSL/TLS context
742 creation, with many options including hostname verification,
743 secure default configuration, lots of documentation and,
744 predefined diffie-hellman keys for perfect forward security
745 and much more. get it while it's still fresh!
746 - use AnyEvent::TLS in AnyEvent::Handle for context management.
747 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
748 so AnyEvent::Socket users have smaller memory footprint
749 in the common case.
750 - add AnyEvent::Handle->push_shutdown method.
751 - add an additional error message parameter to AnyEvent::Handle's
752 on_error callback (for TLS, $! is still available).
753 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
754 - make AnyEvent::Handle more robust against early connection
755 failures (during new), and return C<undef> in that case
756 from the constructor.
757 - AnyEvent::Handle will now try to load only JSON::XS first,
758 then fall back to JSON.
759 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
760 and because it was documented that way.
761
7624.451 Fri Jul 3 00:28:58 CEST 2009
763 - do not clear rbuf when shutting down an AnyEvent::Handle
764 object - doing so breaks AnyEvent::HTTP.
765
7664.45 Mon Jun 29 22:59:26 CEST 2009
767 - a write error could cause AnyEvent::Handle to create
768 an I/O watcher with an undefined $fh.
769 - special-case mapped ipv4 addresses in both
770 AnyEvent::Socket::format_address and parse_address,
771 to treat them just like ipv4 addresses.
772 - updated and overhauled the AnyEvent::Intro doc.
773 - implement AnyEvent::Socket::format_ipv4/ipv6.
774 - slightly speed up portable_pipe/socketpair.
775 - expand condvar begin/end documentation.
776
7774.42 Fri Jun 26 08:32:18 CEST 2009
778 - preliminary and neccesarily incomplete support for IO::Async.
779 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
780 case it was set to IGNORE, to ensure we can catch child statuses
781 even when the calling env acts stupidly.
782 - updated benchmarks with IO::Async, which performs very well.
783
7844.412 Wed Jun 24 01:35:57 CEST 2009
785 - support an "untaint" attribute for AnyEvent::DNS
786 and set it on the default resolver.
787 - implement and document AnyEvent::Impl::Perl::loop.
788 - remove all anyevent-env variables from %ENV
789 when running in tainted mode.
790 - mention and extend the IO::Lambda benchmark.
791
7924.411 Sun Jun 7 18:48:02 CEST 2009
793 - do not try to use F_SETFD on windows.
794
7954.41 Thu May 14 06:40:11 CEST 2009
796 - work around issues in older perls (5.8.0?) when
797 a signal handler is deleted from the %SIG hash.
798 - use POSIX::_exit in child test, to avoid running
799 destructors.
800 - speed up CHLD handling by relying on SIGCHLD being
801 synchronously delivered, even when we roll our own
802 implementation.
803 - AnyEvent::DNS: add the "dname" resource record name
804 for cosmetic reasons.
805
8064.4 Sun Apr 26 20:12:33 CEST 2009
4TODO: - implemented idle watchers, where applicable. 807 - implemented idle watchers, where applicable.
5 - AnyEvent->time died when Event backend was in use. 808 - AnyEvent->time died when Event backend was in use.
6 - fix a memleak in the Tk backend. 809 - fix a memleak in the Tk backend.
7 - sped up Tk timer handling. 810 - sped up Tk timer handling.
811 - clip negative "after" values to 0 in AnyEvent::Impl::Event
812 to avoid spamming out warning messages.
8 - fix Qt timers without interval. 813 - fix Qt timers without interval.
814 - avoid Qt zero-timeout special case, in old libqt's.
9 815
104.352 Mon Apr 20 16:31:11 CEST 2009 8164.352 Mon Apr 20 16:31:11 CEST 2009
11 - fix AnyEvent::Strict error messages for child watchers. 817 - fix AnyEvent::Strict error messages for child watchers.
12 - fix/update Makefile.PL "recommends" versions. 818 - fix/update Makefile.PL "recommends" versions.
13 - add AnyEvent->now_update. 819 - add AnyEvent->now_update.
… …
774.3 Fri Oct 3 09:18:43 CEST 2008 8834.3 Fri Oct 3 09:18:43 CEST 2008
78 - AnyEvent will now install a no-op signal handler for SIGPIPE, 884 - AnyEvent will now install a no-op signal handler for SIGPIPE,
79 unless one has been installed already. 885 unless one has been installed already.
80 - warn about not loaidng AnyEvent::Impl::POE early enough 886 - warn about not loaidng AnyEvent::Impl::POE early enough
81 (patch by Adam Rosenstein). 887 (patch by Adam Rosenstein).
82 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 888 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
83 - fix a bug in where in SSL connect mode, the 889 - fix a bug in where in SSL connect mode, the
84 client would first wait for some data by the server and 890 client would first wait for some data by the server and
85 otherwise hang. (reported and analyzed in an absolutely 891 otherwise hang. (reported and analysed in an absolutely
86 exemplary manner by Adam Rosenstein). 892 exemplary manner by Adam Rosenstein).
87 - fix a bug in where SSL EOF would not be treated 893 - fix a bug in where SSL EOF would not be treated
88 as stream EOF, putting the connection into a hung state. 894 as stream EOF, putting the connection into a hung state.
89 (reported and analyzed in an absolutely exemplary manner by Adam 895 (reported and analysed in an absolutely exemplary manner by Adam
90 Rosenstein). 896 Rosenstein).
91 - fix a potential segfault when the TLS context 897 - fix a potential segfault when the TLS context
92 would go missing in a read callback (e.g. due to stoptls) - 898 would go missing in a read callback (e.g. due to stoptls) -
93 Net::SSLeay of course makes no type checking whatsoever. 899 Net::SSLeay of course makes no type checking whatsoever.
94 - AnyEvent::Handle will not stop the read watcher in TLS mode. 900 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines